Adaptive UAV Navigation Method Based on AHRS.

Area of Science:

• Aerospace Engineering
• Robotics
• Navigation Systems

Background:

• The Constant Acceleration/Constant Velocity (CA/CV) model exhibits inaccuracies in describing UAV relative motion states.
• Accurate relative navigation is crucial for Unmanned Aerial Vehicle (UAV) operations, including formation flying and rendezvous.

Purpose of the Study:

• To develop an adaptive UAV relative navigation method that enhances accuracy by incorporating UAV attitude information.
• To address the limitations of existing state equation models in relative navigation.

Main Methods:

• Utilized Attitude and Heading Reference System (AHRS) output parameters as a benchmark for dead reckoning.
• Derived a relative navigation state equation incorporating attitude error as process noise.
• Integrated an extended Kalman filter with an adaptive decision rule based on filter innovation and employed the Tasmanian Devil Optimization (TDO) algorithm for motion state recalculation.

Main Results:

• The proposed adaptive method significantly reduced relative position errors: 12% (X-axis), 23% (Y-axis), and 32% (Z-axis) compared to the CA/CV model.
• Achieved substantial reductions in relative velocity errors: 350% (X-axis), 330% (Y-axis), and 300% (Z-axis).

Conclusions:

• The adaptive UAV relative navigation method demonstrates superior performance over the CA/CV model.
• The integration of AHRS data and advanced optimization algorithms leads to a significant improvement in relative navigation accuracy for UAVs.